A model is still running

A model is still running or downloading. Leaving will stop the current operation. Stay on this page to let it finish.

Skip to content
LocalAI.toolsBeta

Join the waitlist

Get an email when the new apps and model runs launch.

No account needed.

Huge upgrade coming soon

Think freely.
Run locally.

Powerful AI in your browser or on your desktop.
An infinite private space for your creativity and productivity.

Join for a launch email when the upgraded apps are ready. No account needed.

Local AI, by the numbers

More ways to create, on your own device.

607

Models

Across browser & desktop

Coming soon

Downloads

Available at launch

Coming soon

Local Inferences

Available at launch

Model count includes available browser and desktop models.

esc
F1
F2
F3
F4
F5
F6
F7
F8
F8
F10
F11
F12
~`
!1
@2
#3
$4
%5
^6
&7
*8
(9
)0
—_
+ =
delete
tab
Q
W
E
R
T
Y
U
I
O
P
{[
}]
|\
caps lock
A
S
D
F
G
H
J
K
L
:;
"'
return
shift
Z
X
C
V
B
N
M
<,
>.
?/
shift
fn
control
option
command
command
option
  1. Local inference

    Run AI directly in your browser or on your desktop, powered by your own hardware.

  2. Unlimited runs

    Create, experiment, and try again, without credits or daily run limits.

  3. Absolute Privacy

    Keep your inputs and results on your device, without sending files to cloud AI.

A broad ecosystem of AI

A world of models.
From names you know.

Explore models from well-known AI providers, research teams, and independent creators. Cover a wide range of tasks and use cases.

Models
607
Providers
87
Explore models

Local AI apps

Powerful AI apps.
All running locally.

Create subtitles, edit images and recordings, explore your files, and turn ideas into work—with AI running on your device.

Explore all apps

Build a flow.
Run it locally.

Connect local AI models so the output of one step feeds the next. Start with a preset or build a reusable workflow in the visual editor.

Explore flows
  1. Image inputStart here
    image
  2. Describe imageVision model
    image

    “A quiet lake beneath
    snow-capped mountains.”

    text
  3. Read aloudSpeech model
    textaudio

Download and Run.
Simple as That

  1. 1Find a model
  2. 2Download
  3. 3Run
  • Zero setup hassle

    We handle the model files and runtime setup. No code, dependency installs, or environments to configure.

Local AI.
Made fast.

Put your hardware to work. Batch inputs, run in parallel, and compare models to find your best fit.

One batch. Less busywork.

Add your inputs, run once, and review all your results together.

More done in parallel.

Process several inputs at once with separate model instances, making use of your device’s available capacity.

FLUX.2 Klein sample: red sports car
SD-Turbo sample: red sports car
SD-TurboFLUX.2 Klein

Compare. Then choose.

Use Compare mode to see model results side by side with the same input. Quickly find the model that works best for you.

Built on open foundations.

Frameworks and runtimes that bring AI to your device.

  • Transformers.jsBrowser inference
  • ONNX RuntimeModel execution
  • MediaPipeBrowser vision tasks
  • LiteRT.jsLightweight inference
  • WebAssemblyBrowser CPU execution
  • WebGPUGPU acceleration
  • LiteRT-LMLocal language models
  • CUDADesktop GPU acceleration
  • PyTorchDesktop model execution
  • TransformersDesktop inference
  • MLXApple Silicon inference
  • sherpa-onnxBrowser speaker diarization

A little clarity.
Before you begin.

What to expect when you run AI on your own device.

Read all FAQs
What can I do with LocalAI Tools?

Transcribe speech, remove image backgrounds, translate text, generate speech, and explore other AI tasks. Browse Models to choose a task, then pick a model and a supported configuration. Browser models run on your device.

Do I need an account to use the tools?

You can explore and run browser models without an account. Sign in to use account features and submit issue reports. Local run history is saved in that browser, rather than synced through your account.

Are there generation credits or daily limits?

The available browser tools do not use generation credits or daily generation quotas. Your device’s memory, processing power, storage, and each tool’s input limits determine what you can run.

Are my files uploaded when I run a model?

Browser model inputs are processed locally on your device. Downloading models still requires network requests. Account actions and issue reports use online services; if you attach a file to an issue report, that attachment is uploaded.

What downloads to my browser?

Model weights and the runtime files needed to execute them. The first load may take longer while these files download. Cached files can make later loads faster, but your browser may clear them when storage is low or site data is removed.

Will every model work on my computer or phone?

Compatibility depends on the model, browser, available memory, and hardware. Choose a configuration listed on the model page. Larger models may exceed a phone’s capacity even when the browser supports the required features.

Be first to hear when we launch.